Mycophenolate Mofetil versus Azathioprine for Prevention of Chronic Allograft Dysfunction in Renal Transplantation: The MYSS Follow-Up Randomized, Controlled Clinical Trial
Giuseppe Remuzzi*,,
Paolo Cravedi*,,
Marco Costantini*,
Mariadomenica Lesti*,
Maria Ganeva*,,
Giulia Gherardi*,
Bogdan Ene-Iordache*,
Eliana Gotti,
Donato Donati,
Maurizio Salvadori||,
Silvio Sandrini¶,
Giuseppe Segoloni**,
Stefano Federico,
Paolo Rigotti,
Vito Sparacino,
Piero Ruggenenti*, for the MYSS Follow-Up Study Group
* Clinical Research Center for Rare Diseases, "Aldo e Cele Daccò," Mario Negri Institute for Pharmacological Research, Villa Camozzi, Ranica, and Department of Medicine and Transplantation, Azienda Ospedaliera, Ospedali Riuniti, Bergamo, Italy; Department of Medical Information Services, University Hospital "St. George," Plovdiv, Bulgaria; Unit of Nephrology and Dialysis, Azienda Ospedaliera Universitaria "Ospedale Regionale di Circolo e Fondazione Macchi," Varese, Italy; || Unit of Nephrology and Dialysis, Azienda Ospedaliera Careggi Monna Tessa, Firenze, Italy; ¶ Unit of Nephrology, Dialysis and Transplantation, Azienda Ospedaliera Spedali Civili, Brescia, Italy; ** Unit of Nephrology, Dialysis and Transplantation, Azienda Ospedaliera S.G. Battista, Torino, Italy; Department of Nephrology, Università Federico II, Napoli, Italy; Department of Surgery and Medical Science, Ospedale Giustinianeo, Padova, Italy; and Unit of Renal Transplantation, Ospedale Civico, Palermo, Italy
Address correspondence to: Dr. Piero Ruggenenti, Department of Renal Medicine, "Mario Negri" Institute for Pharmacological Research, Negri Bergamo Laboratories, Via Gavazzeni, 11-24125 Bergamo, Italy. Phone: +39-035-319-888; Fax: +39-035-319-331; E-mail: manuelap{at}marionegri.it
Received for publication October 26, 2006.
Accepted for publication March 16, 2007.
The Mycophenolate Steroids Sparing (MYSS) study found that inrenal transplant recipients who were on immunosuppressive therapywith the cyclosporine microemulsion Neoral, mycophenolate mofetil(MMF) was not better than azathioprine in preventing acute rejectionat 21 mo after transplantation and was 15 times more expensive.The MYSS Follow-up Study, an extension of MYSS, was aimed atcomparing long-term outcome of 248 MYSS patients according totheir original randomization to MMF (1 g twice daily) or azathioprine(75 to 100 mg/d). Primary outcome was estimated GFR at 5 yrafter transplantation. Mean 5-yr GFR difference between azathioprineand mycophenolate was 4.67 ml/min per 1.73 m2 (95% confidenceinterval [CI] –0.43 to 9.77 ml/min per 1.73 m2; P = 0.07).GFR from month 6 (mean ± SEM: 54.3 ± 1.6 versus53.9 ± 1.5 ml/min per 1.73 m2; P = 0.83) to month 72after transplantation (49.5 ± 2.2 versus 47.3 ±2.4 ml/min per 1.73 m2; P = 0.50); GFR slopes (mean ±SEM: –1.10 ± 0.56 versus –1.23 ± 0.31ml/min per 1.73 m2 per year; P = 0.83); and 72-mo patient mortality(4.0 versus 4.0% [P = 0.95]; HR 0.96; 95% CI 0.28 to 3.31; P= 0.95), graft loss (6.8 versus 6.1% [P = 0.82]; HR 0.89; 95%CI 0.32 to 2.46; P = 0.83), incidence of persistent proteinuria(25.0 versus 27.4%; P = 0.72), late (>6 mo after transplantation)rejections (25.3 versus 21.2%; P = 0.53), and adverse eventswere similar on azathioprine (n = 124) and MMF (n = 124), respectively.Outcomes in the two groups were comparable also among patientswith or without steroid therapy, considered separately. In kidneytransplantation, the long-term risk/benefit profile of MMF andazathioprine therapy in combination with cyclosporine Neoralis similar. In view of the cost, standard immunosuppressionregimens for kidney transplantation should perhaps include azathioprinerather than MMF.
Mycophenolate mofetil (MMF), an ester prodrug of mycophenolicacid, was introduced in clinical practice in 1997 as a novelantirejection drug (1). It suppresses proliferation of T andB lymphocytes by inhibiting inosine monophosphate dehydrogenase,a crucial enzyme in the de novo pathway of purine synthesisin the S phase of the cell cycle (2). In other eukaryotic cells,however, inosine monophosphate dehydrogenase inhibition haslittle effect on cell division because purines can also be generatedfrom nucleotide breakdown products. Because lymphocytes lackthis "salvage pathway," MMF specifically inhibits their proliferation,thereby limiting cell-mediated immunity, but has little impacton other tissues with high proliferative activity, such as skin,intestine, and bone marrow (2). This property has been presentedas a major advantage over other, less selective antiproliferativeagents such as azathioprine (3).
Three groundbreaking, prospective, randomized, double-blindclinical trials in a total of 1593 patients who received animmunosuppressive regimen that included cyclosporine Sandimmune(Novartis, Basel, Switzerland) and steroids (4–6) founda significantly lower proportion of patients with a first rejectionepisode on MMF 2 g/d (33.2%) or 3 g/d (35.0%) than on azathioprineor placebo (50.3%). A pooled analysis of the three trials foundthat this trend was still consistent at 1 yr (7). On the basisof these findings, European regulatory agencies and the US Foodand Drug Administration approved MMF for the prevention of rejectionin renal transplantation. In the subsequent years, MMF progressivelyreplaced azathioprine as a part of standard treatment for preventingrejection in solid organ and bone marrow transplantation inmost centers worldwide (8).
In 1997, however, at the same time MMF was introduced in clinicalpractice, a microemulsion preparation of cyclosporine, Neoral(Novartis), became available. The microemulsion was developedto provide greater and more consistent exposure to cyclosporinethan the older, oil-based Sandimmune formulation (9). This facilitatedthe individual tailoring and monitoring of cyclosporine therapy,which translated into a reduced incidence of acute rejectionepisodes and an improved graft survival, without increasingadverse effects related to cyclosporine toxicity, not only inkidney transplantation (10,11) but also in other clinical settingssuch as heart (12) and liver (13) transplantation. As a resultof the better risk/benefit profile, Neoral soon became the preferredform of cyclosporine in many centers (9,11,14).
The Mycophenolate Steroids Sparing (MYSS) study published in2004 (15) was designed to test formally whether MMF retainsits better antirejection activity even in immunosuppressiveregimens that use the cyclosporine microemulsion formulationNeoral as opposed to Sandimmune. The MYSS trial was a fullyindependent, academic study and was not funded by pharmaceuticalcompanies. It compared the incidence of acute rejections in336 patients who were randomly assigned to MMF or azathioprinein the context of a Neoral-based immunosuppressive regimen.Data showed that MMF was not more effective than azathioprinein preventing acute rejection at 21 mo after transplantation,had a similar tolerability profile, but was 15 times more expensive.Because of the relatively limited follow-up, however, this studycould not assess the effects of MMF and azathioprine on theonset and progression of chronic allograft dysfunction (16),a syndrome of proteinuria and worsening renal function withprogressive nephron loss and scarring of the graft (chronicallograft nephropathy). This is a key issue because, after recipientdeath, chronic allograft dysfunction represents the major causeof graft loss in the long term (17). Moreover, results fromregistry analyses showed that continued treatment with MMF versusazathioprine was associated with a protective effect againstrenal function deterioration beyond 1 yr after transplantation(18) and superior graft survival at 4 yr (19). Despite the limitationsof the retrospective design of these analyses, these data furtherlimited the possibility to use results of the MYSS trial tochange the practice of most transplant centers to regard MMFas a key component of immunosuppressive drug regimens that arebased on cyclosporine microemulsion.
To address this issue, we designed the MYSS Follow-up study.This was an extension of the MYSS study and prospectively comparedlong-term outcomes of the two cohorts of MYSS patients in thesetting of a similar immunosuppressive regimen based on themicroemulsion Neoral, according to their original randomizationto MMF or azathioprine. The results of this MYSS Follow-up studyformed the basis of this report.
Patients and Study Design
The MYSS study was a multicenter, prospective, randomized, parallel-grouptrial that compared the incidence of acute rejection in recipientsof a first kidney transplant from deceased donors who were randomlyassigned on a 1:1 basis within each center to receive treatmentwith 1 g of MMF twice daily or azathioprine once daily (100mg if body weight <75 kg, 150 mg if 75 kg) in combinationwith maintenance immunosuppressive therapy with the cyclosporinemicroemulsion Neoral and steroids (see Remuzzi et al. [15] forfurther details). From October 1997 to May 2001, 336 patientswere included. However, two patients (one per group) did notreceive the transplant for technical reasons. Thus, 334 patients(167 per group) received the study drugs (Figure 1). The randomizationwas centralized at the Laboratory of Biostatistics of the ClinicalResearch Centre for Rare Diseases of the Mario Negri Institutefor Pharmacologic Research. At completion of the first 6-motreatment period (phase A), steroid dosage was progressivelytapered and discontinued in patients with (1) no more than twoacute rejection episodes, (2) no steroid-resistant rejections,and (3) serum creatinine concentrations of 177 µmol/L(Figure 1). When an acute rejection episode was diagnosed, oralsteroid was renewed at the dosage before the last reductionor at 6 mg every other day if the patient had already discontinuedthe medication. Patients were then followed up to 21 mo aftertransplantation (phase B).
Fifty-three patients (28 on azathioprine, 25 on MMF) who wereincluded in the MYSS core trial and referred to the Centersof Genoa, Antwerp, and Montpellier were not included becauseof fund restrictions that did not allow covering the costs foractive patient follow-up and data monitoring and recording.Of the remaining 281 patients who were included in the MYSStrial and referred to the Bergamo, Padua, Brescia, Turin, Varese,Florence, Palermo, and Naples transplant centers, 248 patientsentered the MYSS Follow-up study, whereas 33 did not. Reasonsfor exclusion were death (n = 2, both on azathioprine), dialysis(n = 2 in each group), and withdrawal of consent (n = 27, 11in the azathioprine, 16 in the MMF group). Donor and baselinecharacteristics, immunosuppressive therapy, and concomitantmedications of patients who were or were not included in thefollow-up study, as well as of patients who did or did not completethe study, were similar (data not shown).
Patients were prospectively followed up to October 2004 by visitsevery 6 mo (±15 d). Clinical and laboratory parameterswere recorded in dedicated forms and were doubly entered. Finally,MYSS and MYSS Follow-up databases were merged in a unique databasethat served for final analyses. All of these activities werecoordinated and monitored by the Clinical Research Centre ofthe Mario Negri Institute without the involvement of pharmaceuticalindustry sponsorship.
All patients provided written informed consent to enter theMYSS study and a subsequent oral consent to have their follow-upextended for at least 2 yr. Outcome data were recorded withoutinterfering with patient routine clinical management and werehandled according to standard regulations for data registration,use, and preservation of patient anonymity and privacy.
Outcomes
Primary outcome variable was estimated GFR (eGFR). Secondaryoutcome variables were the rate of GFR decline over time (GFRslope), the incidence of new-onset persistent proteinuria (urinaryprotein excretion >0.5 g/24 h in two consecutive visits inpatients without residual proteinuria since transplantation),patient and graft survival, acute rejections, and adverse events.
BP, serum creatinine concentration, proteinuria and other relevantlaboratory parameters, treatments, and adverse events includingacute rejections were recorded every 6 mo up to study end. TheGFR was estimated on the basis of the Walser equation (men:GFR = 7.57 x (serum creatinine x 0.0884)–1 – 0.103x age + 0.096 x weight – 6.66; women: GFR = 6.06 x (serumcreatinine x 0.0884)–1 – 0.080 x age + 0.080 x weight– 4.81), the equation that in renal transplantation generatesthe GFR estimates that better predict true GFR as measured bythe iohexol plasma clearance (20). Complementary analyses usedGFR data as estimated by the Modification of Diet in Renal Disease(MDRD) and Nankivell equations (21,22). The GFR slope was estimatedin patients who had at least two GFR measurements from month6 after transplantation (taken as baseline) to study end. Thediagnosis of acute rejection episodes was made on the basisof clinical criteria as in the MYSS study (15). Kidney biopsysamples were taken whenever appropriate to confirm the diagnosisand for all steroid-resistant rejection episodes.
Statistical Analyses
All analyses were performed according to the original randomizationin the MYSS trial. An additional "on treatment" analysis wasperformed in patients who assumed only the study drug as perprotocol. Analyses first considered the study population asa whole, and then patients who completed steroid withdrawalwhile in the MYSS study separately from those who did not. Foravoiding the confounding effect of different follow-up periods,main outcomes were also described in three subcohorts of patientswho were actively followed for at least 24, 48, and 72 mo, respectively.
Baseline recipient and donor characteristics were compared accordingto treatment groups by means of t test, 2, or Fisher exact testas appropriate. Kaplan-Meier analysis was used to investigatewhether survival functions differed between treatment groups.The difference between the survival curves was assessed by log-ranktest. Cox proportional hazards models, with adjustment for maindemographic and clinical covariates (site, donor and recipientgender, donor and recipient age, donor and recipient weight,and donor serum creatinine) were used to detect the effect oftreatment on time-dependent end points. Proportional hazardsassumptions were checked by means of Schoenfeld residuals. Generallinear mixed models were used to compare GFR levels at varioustime points (e.g., every 6 mo since transplantation, betweenregimen treatments). Statistical analyses were accomplishedwith SAS software (version 9.1.3; SAS Institute, Cary, NC).The statistical level of significance was P < 0.05. All statisticalanalyses were performed by two-tailed tests. The study was notinitially powered to detect a difference in graft survival orfunction between treatment groups, and the recruitment was finalizedto have all MYSS patients on active follow-up for at least 2yr in the setting of the MYSS Follow-up study.
Patients
Main donor and recipient characteristics at time of transplantationwere similar in the two treatment groups—except for asignificantly higher percentage of female patients in the azathioprinearm—in the study population as a whole as well as in thetwo cohorts of patients who did or did not complete the steroidwithdrawal protocol (Table 1). Cold ischemia time was similarbetween the two treatment arms in the study group as whole (azathioprine16.6 ± 7.2 h; MMF 16.0 ± 6.5 h), as well as inthe two subgroups that did (azathioprine 15.6 ± 6.7 h,MMF 15.7 ± 6.1 h) or did not complete (azathioprine 19.2± 7.9 h; MMF 16.7 ± 7.1 h) steroid withdrawal.Patients were followed for a median (interquartile range) of65.5 mo (52.7 to 73.6) in the azathioprine group and of 64.1mo (53.8 to 73.6) in the MMF group. The flowchart of the studyis shown in Figure 1. At last follow-up, 18.5% of patients whowere randomly assigned to azathioprine and 8.1% of those whowere randomly assigned to MMF were on reduced dosages of thestudy drugs compared with the dosages at 6 mo after transplantation(i.e., before steroid reduction or withdrawal).
Table 1. Baseline characteristics of recipients and donorsa
Primary Efficacy Outcome GFR.
All 248 patients had at least one GFR evaluated during the follow-up.Throughout the whole study period, eGFR was similar in the twotreatment groups at each 6-mo visit (Figure 2). Mean GFR differencebetween azathioprine and MMF at 60 mo from transplantation was4.67 ml/min per 1.73 m2 (95% confidence interval [CI] –0.43to 9.77 ml/min per 1.73 m2; P = 0.07; Table 2). At month 72after randomization, mean (±SE) GFR was 49.5 ±2.2 ml/min per 1.73 m2 in the azathioprine group and 47.3 ±2.4 ml/min per 1.73 m2 in the MMF group (P = 0.50). eGFR valuesat various time points after transplantation were similar inthe two treatment groups even within cohorts of patients withthe same follow-up duration (Figure 3). Consistent results wereobtained by "on treatment" analyses (data not shown). Amongpatients who completed steroid withdrawal considered as a whole,the GFR progressively declined over time with a similar trendin both treatment groups (Figure 4, top). At variance, amongpatients who did not complete steroid withdrawal, those whowere on azathioprine had a GFR with a slight progressive increaseduring the whole follow-up, whereas those who were on MMF hada slowly declining GFR (Figure 4, bottom). This different trendresulted in a numerically higher GFR at 72 mo after randomizationin patients who were on azathioprine (51.2 ± 4.8 ml/minper 1.73 m2) than in those who were on MMF (36.2 ± 4.2ml/min per 1.73 m2; P = 0.07).
Figure 2. Estimated GFR (eGFR) values at various time points from transplantation in the two treatment groups according to the original patient randomization in the Mycophenolate Steroids Sparing (MYSS) study. Error bars indicate SE. *P < 0.05 versus mycophenolate mofetil (MMF).
Figure 3. eGFR values at various time points from transplantation in cohorts with at least 24 (left; azathioprine n = 111, MMF n = 116), 48 (middle; azathioprine n = 94, MMF n = 99), and 72 (right; azathioprine n = 40, MMF n = 41) months of follow-up. Error bars indicate SE.
Figure 4. eGFR values at various time points from transplantation according to treatment groups in patients who did (top) or did not complete (bottom) steroid withdrawal. Error bars indicate SE. *P < 0.05 versus MMF.
Secondary Efficacy Outcomes Patient and Graft Outcome.
The 72-mo patient (azathioprine 4.0%; MMF 4.0%; P = 0.95) andgraft loss (azathioprine 6.8%; MMF 6.1%; P = 0.83) were virtuallyidentical in the two treatment groups (Figure 5). Patient mortalitywas also similar in the two cohorts of patients who did (azathioprine2.5%; MMF 2.6%; P = 0.96) or did not complete (azathioprine7.0%; MMF 6.2%; P = 0.80) steroid withdrawal considered separately,as well as graft loss (completing steroid withdrawal: azathioprine3.8%; MMF 2.8%; P = 0.71; not completing steroid withdrawal:azathioprine 11.9%; MMF 11.1%; P = 0.94). Cox model did notshow any difference between treatment groups even when adjustedfor demographic and clinical covariates, cold ischemia time,steroid withdrawal, and azathioprine or MMF dosage reduction(at last follow up versus month 6 after transplantation; datanot shown).
Figure 5. Cumulative patient mortality (top) and graft loss (bottom) probability in the two treatment groups according to the original patient randomization.
GFR Slopes.
Data for slope analyses were available for 121 patients whowere on azathioprine and 124 who were on MMF. GFR slopes weresimilar in the two treatment arms (azathioprine –1.10± 0.56 ml/min per 1.73 m2 per year; MMF –1.23 ±0.31 ml/min per 1.73 m2 per year; P = 0.83) in the study groupas a whole, as well as in the subgroups that did (azathioprine–1.73 ± 0.39 ml/min per 1.73 m2 per year; MMF –1.24± 0.38 ml/min per 1.73 m2 per year; P = 0.36) or didnot complete (azathioprine 0.19 ± 1.49 ml/min per 1.73m2 per year; MMF –1.23 ± 0.55 ml/min per 1.73 m2per year; P = 0.35) steroid withdrawal.
Proteinuria.
During the whole follow-up period, 31 (25.0%) patients in theazathioprine group and 34 (27.4%) patients in the MMF groupdeveloped persistent clinical proteinuria (P = 0.67). The numberof patients who developed persistent proteinuria was similarin the two treatment groups even among patients who did (azathioprinen = 21 [25.9%]; MMF n = 20 [26.3%]) or did not complete (azathioprinen = 10 [23.3%]; MMF n = 14 [29.2%] steroid withdrawal.
Adverse Events
Sixty-five (52.4%) patients in the azathioprine arm and 57 (46.0%)in the MMF arm experienced at least one acute rejection episodethroughout the whole study period (P = 0.31; Figure 6). Thenumber of patients who had late occurrence of rejection (>6mo after transplantation) and were on azathioprine (n = 20,25.3%) and on MMF (n = 18, 21.2%) was virtually identical aswell (P = 0.53). Among patients who completed steroid withdrawal,the numbers of patients with acute rejections were similar inthe two study groups (azathioprine n = 40 [49.4%]; MMF n = 31[40.8%]; P = 0.28). The trend was similar in those who did notcomplete steroid withdrawal (azathioprine n = 25 [58.1%]; MMFn = 26 (54.2%); P = 0.70).
Figure 6. Cumulative probability of acute rejection episodes in the two treatment groups according to the original patient randomization.
The total number of infections was similar in the two studygroups, with no significant differences in the incidence ofcytomegalovirus reactivations (Table 3). The number of cardiovascularevents during the whole follow-up period was similar in thetwo treatment arms, whereas there was a NS trend to more neoplasmsin those who were on azathioprine (Table 3). The overall incidenceof adverse events was similar in the two treatment groups andalso in the two cohorts of patients who did or did not completesteroid withdrawal considered separately (Table 3).
Immunosuppressive Therapy and Concomitant Medications
Cyclosporine trough levels throughout the whole study periodwere similar in the two treatment groups in both cohorts ofpatients who did or did not complete steroid withdrawal (Figure 7).The average dosages of azathioprine (ranging from 69 to 94 mg/d)and MMF (1533 to 1804 mg/d) at each time point of the follow-upwere similar between cohorts of patients who did or did notcomplete steroid withdrawal. Among patients who did not completesteroid withdrawal, the average dosages of steroid (4 to 6 mg/d)were similar in the two treatment groups. Throughout the studyperiod, the proportions of patients who received at least oncean angiotensin-converting enzyme inhibitor (35.5 versus 36.3%),angiotensin II receptor antagonist (2.4 versus 4.0%), statin(38.7 versus 29.8%), or aspirin (43.6 versus 50.0%) therapyin the azathioprine and MMF groups, respectively, were similar.
Figure 7. Cyclosporine trough levels at various time points from transplantation in the two treatment groups in patients who did (top) or did not complete (bottom) steroid withdrawal. Error bars indicate SE.
The MYSS Follow-up study was the longest prospective, randomizedclinical trial ever attempted in kidney transplantation to compareformally the effects of various immunosuppressive drug regimenson patient and graft outcomes. Post hoc analyses of previousregistration trials of MMF in kidney transplantation extendedthe observation period to a maximum of 3 yr after randomization.In the MYSS Follow-up study, outcome data were recorded overa median of 5.5 yr after patient allocation to MMF or azathioprinetreatment. Analyses that were performed according to the originalrandomization in MYSS or to the actual treatment assumed onfollow-up consistently showed that MMF was no better than azathioprinein preventing chronic allograft dysfunction in deceased-donorrenal transplant recipients who were on a cyclosporine Neoral-basedimmunosuppressive therapy. eGFR was similar in the two groupsat 60 mo after transplantation—the primary outcome ofthe trial—as well as at each study visit. Consistently,GFR slopes; patient and graft survival; and the incidence ofacute allograft rejections, new-onset proteinuria, and adverseevents were similar in both groups throughout the whole studyperiod. The outcome was similar in the two treatment arms evenwithin the two subgroups that did or did not complete steroidwithdrawal.
MMF and azathioprine were also equally well tolerated. Altogether,data of the MYSS Follow-up study suggest that MMF does not havea long-term better risk-benefit profile than azathioprine, evenin the setting of a dual immunosuppressive regimen that doesnot include oral steroids. These findings are in harmony withdata from the registration trials that showed a similar 3-yrmortality and graft loss in patients who were on azathioprine(23,24) or placebo (25) compared with those who were on MMF.Of note, however, regardless of treatment randomization, patientor graft loss was remarkably higher in previous trials (on average,19.9% at 3 yr) than in our study (10.4% over 5.5 yr). Worselong-term outcome was also associated with a higher incidenceof acute rejections at 6 mo (on average, 42.5%) than in ourstudy (34.5%). This was not explained by differences in thestudy treatments, because the daily dosages of azathioprine(100 or 150 mg if body weight < or 75 kg, respectively) andMMF (2 g) were similar. Therefore, concomitant treatment mostlikely had a role. All patients who were included in the registrationtrials were on long-term steroid therapy, whereas a substantialproportion of patients who were included in the MYSS Follow-upstudy withdrew steroids starting from month 6 after transplantation.Therefore, despite the seemingly less intensive concomitantimmunosuppressive regimen, acute rejections were less frequentand long-term patient and graft survival were higher in theMYSS Follow-up study than in the registration trials. This maybe the consequence of time-dependent factors that resulted inbetter patient care in more recent studies. However, anotherplausible—even if not exclusive—explanation maybe the introduction of the cyclosporine microemulsion Neoralthat was used in all patients who were included in the MYSSFollow-up study, but that was not available when the registrationtrials were performed. Similar confounders may explain why previousretrospective registry analyses (18,19) found a worse long-termpatient and graft survival in patients who were on azathioprine,who in the large majority of cases received a kidney transplantbefore 1997 (i.e., before Neoral became available), than inthose who were on MMF and received a graft in the followingyears and, therefore, in the large majority of cases were onconcomitant treatment with Neoral. Consistent with this interpretationare data from two studies that were not biased by the confoundingeffect of time or concomitant medications: A pooled analysisof the three pivotal registration trials showing no differencesin 3-yr graft function and survival on MMF as compared withazathioprine or placebo (26), and a paired kidney analysis showingsimilar graft function and survival in 476 renal transplantrecipients who were on continued MMF or azathioprine therapyduring a mean follow-up of 3.3 yr (27). Altogether, these datacombined with MYSS Follow-up study findings, challenge the commonbelief that MMF has a specific protective effect against thedevelopment of chronic allograft nephropathy (CAN). This possibilitywas suggested by experimental data that MMF, in addition toinhibiting lymphocyte proliferation, has an antiproliferativeeffect on mesangial and vascular smooth muscle cells, monocytes,and fibroblasts that might in theory prevent or limit the underlyingscarring processes of CAN. Other potentially beneficial effectsof MMF include inhibition of adhesion molecule expression onendothelial cells—which may retard tissue infiltrationby inflammatory cells—increased apoptosis of monocytes,and reduced antibody production from B cells (reviewed in reference[28]). That these effects may translate into a clinically relevantbenefit in humans, however, is unproved. Data in support ofthis possibility are generated by sequential studies showingimprovements in graft structure and function after replacementof calcineurin inhibitors with MMF (29,30). These findings,however, may reflect a spontaneous recovery from the nephrotoxiceffects of previous cyclosporine or tacrolimus treatment, ratherthan a specific protective effect of MMF against CAN. Observationalstudies showing a lower incidence of chronic interstitial fibrosiswith MMF than with azathioprine therapy were likely affectedby the confounding effect of time, because outcome data werenot analyzed in the setting of a randomized, prospective design(31). Only one randomized clinical trial prospectively comparedthe effects of MMF and azathioprine on the incidence of biopsy-provenCAN (32). On the basis of the analysis of the Banff score, MMFas compared with azathioprine therapy was associated with alower incidence of nephropathy at 1 yr after transplantation.However, closer examination of the data (33) revealed that allograftglomerulopathy, mesangial matrix increase, vascular changes,interstitial fibrosis, and tubular atrophy, as well as renalfunction, were similar in the two treatment groups. These findingsdid not allow the conclusion that MMF is superior to azathioprinein preventing CAN (33). Actually, in the long term, the oppositemight be true, as suggested by a recent observational studyof 1511 renal transplant patients that showed that at 8.5 yrafter transplantation, those who were on MMF had significantlylower GFR than those who were on azathioprine (34).
Altogether, the MYSS and MYSS Follow-up studies consistentlyshow that MMF and azathioprine are equally effective in preventingboth acute rejection and CAN (an equivalent of chronic rejection)in kidney transplantation. Studies suggest that MMF may be nobetter than azathioprine also in other areas of organ transplantation.A randomized clinical trial recently found that MMF and azathioprinewere equally effective in preventing acute rejection and obliterativebronchiolitis (the pathologic entity that represents "chronicrejection" in pulmonary allografts) in lung transplant recipientswho were on immunosuppressive treatment with cyclosporine Neoraland corticosteroids (35).
Articles that were published after the registration studiesarose also some concern about the safety profile of MMF comparedwith azathioprine. A meta-analysis of 6387 renal transplantpatients who were included in 20 randomized clinical trialsof MMF versus azathioprine (36) found a significantly higherincidence of diarrhea, leucopenia, and cytomegalovirus diseasein those who were on MMF. The Transplant European Survey onAnemia Management (37), a 5-yr survey of 4263 renal transplantrecipients, showed significantly lower hemoglobin in patientswho were on any immunosuppressive drug combination that includedMMF, a finding that was at least in part explained by lowerGFR. Recent reports showed also an increased risk for tuberculosisin renal transplant patients who received an immunosuppressiveregimen that included MMF rather than azathioprine (38). However,in our study, we found a NS trend to more neoplasms on azathioprinethan on MMF. Although the study was not powered to assess theimpact of treatment on these outcomes and a random effect couldnot be definitely ruled out, these findings are in harmony withresults of previous retrospective analyses that showed a higherincidence of skin and nondermatologic malignancies with azathioprinethan with MMF therapy (39).
The MYSS Follow-up study has some limitations. Long-term outcomewas remarkably good regardless of treatment randomization, afinding that may reflect the effectiveness of both immunosuppressiveregimens under evaluation but also the inclusion of relativelylow-risk patients, such as relatively young white patients whowere at their first kidney transplant and who seldom had diabetesas a primary cause of ESRD. Data from registration trials (23–26)and observational studies (27), however, suggest that azathioprineand MMF may have a similar long-term risk/benefit profile alsoin second transplant recipients and in patients who are at higherimmunologic risk and are on concomitant treatment with cyclosporine(or tacrolimus) and steroid, regardless of Thymoglobulin inductiontherapy. Whether MMF and azathioprine are equally effectivealso in renal transplant patients who are on tacrolimus-basedimmunosuppressive regimens remains to be addressed. Also, arelatively small number of patients were available for comparativeanalyses after >5 yr of follow-up, and, as in previous registrationtrials (23–25), the sample size was not estimated a priorion the basis of an expected effect on long-term efficacy variables.A confounding effect of a survival bias on GFR analyses wasunlikely because patient and kidney survival were similar inthe two treatment groups, and GFR and event analyses were consistentin showing similar long-term outcomes on MMF or azathioprine.Moreover, similar GFR outcomes were observed also when the analyseswere performed in subgroups of patients with homogeneous follow-up.Finally, failure to detect a long-term benefit of MMF over azathioprinewas not explained by a confounding effect of a reduction inthe MMF dosage aimed to limit the risk of thrombocytopenia orleukopenia after steroid withdrawal. Indeed, at last follow-up,the proportion of patients who were on reduced dosages of MMFwas consistently less than the proportion of patients who wereon reduced dosages of azathioprine, and, at multivariate analyses,patient and graft outcomes were similar in the two treatmentgroups even after adjustment for dosage reduction.
Data from our study, consistent with post hoc analyses of previousregistration trials (23–26) and recent observational studies(27), suggest that the long-term risk/benefit profile of MMFand azathioprine therapy in kidney transplantation are similar,even in the setting of a cyclosporine-based immunosuppressiveregimen that does not include steroids. Because the costs forstandard treatment with MMF remarkably exceed those of azathioprine(15), standard immunosuppression regimens for kidney transplantationshould perhaps include azathioprine rather than MMF.
Marco Costantini is a recipient of a fellowship from AssociazioneRicerca Trapianti, Milan, Italy.
Organization of the MYSS Follow-up Study Group: Principal investigator,Giuseppe Remuzzi, Azienda Ospedaliera, Ospedali Riuniti, Bergamo,Istituto di Ricerche Farmacologiche Mario Negri; study coordinator,Piero Ruggenenti, Azienda Ospedaliera, Ospedali Riuniti, Bergamo,Istituto di Ricerche Farmacologiche Mario Negri. Ethical Committee:Livio Robba, Francesco Vaccari, Ottavio Roberto, Luigi Minetti,Valentina Paris, Eugenio Battaglia; Steering Committee: GiuseppeRemuzzi, Norberto Perico, Piero Ruggenenti, Silvio Sandrini,Giuseppe Segoloni. Investigators and Institutions (the numberof included patients is in parentheses): Azienda Ospedaliera,Ospedali Riuniti, Bergamo (n = 89): Principal Investigator,Eliana Gotti; Co-Investigator, Giuseppe Locatelli, Piero Ruggenenti,Giovanni Rota; Ospedale Regionale di Circolo e Fondazione Macchi,Varese (n = 47): Principal Investigator, Luigi Gastaldi; Co-Investigator,Donato Donati; Azienda Ospedaliera Careggi Monna Tessa, Firenze(n = 44): Principal Investigator, Maurizio Salvadori; Co-Investigator,Rosa Piperno, Elisabetta Bertoni; Azienda Ospedaliera SpedaliCivili, Brescia (n = 42): Principal Investigator, Rosario Maiorca;Co-Investigator, Silvio Sandrini, Gisella Setti; Azienda OspedalieraS.G. Battista, Torino (n = 27): Principal Investigator, GiuseppeSegoloni; Co-Investigator, Giuseppe Piccoli; UniversitàFederico II, Napoli (n = 12): Principal Investigator, StefanoFederico; Ospedale Giustinianeo, Padova (n = 12): PrincipalInvestigator, Paolo Rigotti; Co-Investigator, Ermanno Ancona,Nicola Baldan; Ospedale Civico, Palermo (n = 10): PrincipalInvestigator, Vito Sparacino; Co-Investigator, Sergio Calabrese.Laboratory measurements: F. Gaspari, D. Cattaneo, R. Caruso,S. Merlini, S. Baldelli, S. Ferrari, N. Stucchi, E. Centemeri,S. Zenoni, F. Carrara, M. Pellegrino (Istituto di Ricerche FarmacologicheMario Negri); study monitoring: M. Lesti, G. Gherardi (Istitutodi Ricerche Farmacologiche Mario Negri); statistical analysis:M. Costantini, M. Ganeva, A. Perna (Istituto di Ricerche FarmacologicheMario Negri); data management: B. Ene-Iordache (Istituto diRicerche Farmacologiche Mario Negri).
We are grateful to Manuela Vergani, Franca Gamba, Emilia Camoni,and all of the nurses of the participating centers for precioushelp in patient care and follow-up. Annalisa Perna helped toprepare the study protocol and to perform the analyses; ManuelaPassera helped to prepare the manuscript.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
Fulton B, Markham A: Mycophenolate mofetil. A review of its pharmacodynamic and pharmacokinetic properties and clinical efficacy in renal transplantation.
Drugs 51
: 278
–298, 1996[Medline]
Eugui EM, Almquist SJ, Muller CD, Allison AC: Lymphocyte-selective cytostatic and immunosuppressive effects of mycophenolic acid in vitro: Role of deoxyguanosine nucleotide depletion.
Scand J Immunol 33
: 161
–173, 1991[CrossRef][Medline]
Platz KP, Sollinger HW, Hullett DA, Eckhoff DE, Eugui EM, Allison AC: RS-61443, a new, potent immunosuppressive agent.
Transplantation 51
: 27
–31, 1991[Medline]
Sollinger HW; for the US Renal Transplant Mycophenolate Mofetil Study Group: Mycophenolate mofetil for the prevention of acute rejection in primary cadaveric renal allograft recipients.
Transplantation 60
: 225
–232, 1995[Medline]
The Tricontinental Mycophenolate Mofetil Renal Transplantation Study Group: A blinded, randomized clinical trial of mycophenolate mofetil for the prevention of acute rejection in cadaveric renal transplantation.
Transplantation 61
: 1029
–1037, 1996[CrossRef][Medline]
European Mycophenolate Mofetil Cooperative Study Group: Placebo-controlled study of mycophenolate mofetil combined with cyclosporin and corticosteroids for prevention of acute rejection.
Lancet 345
: 1321
–1325, 1995[Medline]
Halloran P, Mathew T, Tomlanovich S, Groth C, Hooftman L, Barker C: Mycophenolate mofetil in renal allograft recipients: A pooled efficacy analysis of the three randomized, double-blind, clinical studies in prevention of rejection. The International Mycophenolate Mofetil Renal Transplant Study Groups.
Transplantation 63
: 39
–47, 1997
[published erratum appears in Transplantation 63: 618, 1997][CrossRef][Medline]
Mele TS, Halloran PF: The use of mycophenolate mofetil in transplant recipients.
Immunopharmacology 47
: 215
–245, 2000[CrossRef][Medline]
Keown P, Niese D; on behalf of the International Sandimmune Neoral Study Group: Cyclosporine microemulsion increases drug exposure and reduces acute rejection without incremental toxicity in de novo renal transplantation.
Kidney Int 54
: 938
–944, 1998[CrossRef][Medline]
Shah MB, Martin JE, Schroeder TJ, First MR: The evaluation of the safety and tolerability of two formulations of cyclosporine: NEORAL and Sandimmune. A meta-analysis.
Transplantation 67
: 1411
–1417, 1999[CrossRef][Medline]
Pollard SG, Lear PA, Ready AR, Moore RH, Johnson RW: Comparison of microemulsion and conventional formulations of cyclosporine A in preventing acute rejection in de novo kidney transplant patients. The UK NEORAL Renal Study Group.
Transplantation 68
: 1325
–1331, 1999[CrossRef][Medline]
Eisen HJ, Hobbs RE, Davis SF, Carrier M, Mancini DM, Smith A, Valantine H, Ventura H, Mehra M, Vachiery JL, Rayburn BK, Canver CC, Laufer G, Costanzo MR, Copeland J, Dureau G, Frazier OH, Dorent R, Hauptman PJ, Kells C, Masters R, Michaud JL, Paradis I, Renlund DG, Vanhaecke J, Mellein B, Mueller EA: Safety, tolerability, and efficacy of cyclosporine microemulsion in heart transplant recipients: A randomized, multicenter, double-blind comparison with the oil-based formulation of cyclosporine—Results at 24 months after transplantation.
Transplantation 71
: 70
–78, 2001[CrossRef][Medline]
Graziadei IW, Wiesner RH, Marotta PJ, Porayko MK, Dalke LJ, Wilson SM, Steers J, Krom RA: Neoral compared to Sandimmune is associated with a decrease in histologic severity of rejection in patients undergoing primary liver transplantation.
Transplantation 64
: 726
–731, 1997[CrossRef][Medline]
Gaspari F, Perico N, Pisoni R, Anedda MF, Signorini O, Remuzzi G: How to convert from traditional cyclosporine to the microemulsion formulation in stable renal transplant patients?
Clin Transplant 12
: 379
–390, 1998[Medline]
Remuzzi G, Lesti M, Gotti E, Ganeva M, Dimitrov BD, Ene-Iordache B, Gherardi G, Donati D, Salvadori M, Sandrini S, Valente U, Segoloni G, Mourad G, Federico S, Rigotti P, Sparacino V, Bosmans JL, Perico N, Ruggenenti P: Mycophenolate mofetil versus azathioprine for prevention of acute rejection in renal transplantation (MYSS): A randomized trial.
Lancet 364
: 503
–512, 2004[CrossRef][Medline]
Maes B, Vanrenterghem Y: MYSS: Should we alter clinical practice?
Lancet 364
: 2016
–2017, 2004[Medline]
Pascual M, Theruvath T, Kawai T, Tolkoff-Rubin N, Cosimi B: Strategies to improve long-term outcomes after renal transplantation.
N Engl J Med 346
: 580
–590, 2002[Free Full Text]
Meier-Kriesche HU, Steffen BJ, Hochberg AM, Gordon RD, Liebman MN, Morris JA, Kaplan B: Mycophenolate mofetil versus azathioprine therapy is associated with a significant protection against long-term renal allograft function deterioration.
Transplantation 75
: 1341
–1346, 2003[CrossRef][Medline]
Ojo AO, Meier-Kriesche HU, Hanson JA, Leichtman AB, Cibrik D, Magee JC, Wolfe RA, Agodoa LY, Kaplan B: Mycophenolate mofetil reduces late renal allograft loss independent of acute rejection.
Transplantation 69
: 2405
–2409, 2000[CrossRef][Medline]
Gaspari F, Ferrari S, Stucchi N, Centemeri E, Carrara F, Pellegrino M, Gherardi G, Gotti E, Segoloni G, Salvadori M, Rigotti P, Valente U, Donati D, Sandrini S, Sparacino V, Remuzzi G, Perico N; MYSS Study Investigators: Performance of different prediction equations for estimating renal function in kidney transplantation.
Am J Transplant 4
: 1826
–1835, 2004[CrossRef][Medline]
Levey AS, Greene T, Kusek JW, Beck GJ; for the MDRD Study Group: A simplified equation to predict glomerular filtration rate from serum creatinine [Abstract].
J Am Soc Nephrol 11
: 155A
, 2000
US Renal Transplant Mycophenolate Mofetil Study Group: Mycophenolate mofetil in cadaveric renal transplantation.
Am J Kidney Dis 34
: 296
–303, 1999[Medline]
European Mycophenolate Mofetil Cooperative Study Group: Mycophenolate mofetil in renal transplantation: 3-year results from the placebo-controlled trial.
Transplantation 68
: 391
–396, 1999[CrossRef][Medline]
Mathew TH: A blinded, long-term, randomized, multicenter study of mycophenolate mofetil in cadaveric renal transplantation: Results at 3 years. Tricontinental Mycophenolate Mofetil Renal Transplantation Study Group.
Transplantation 65
: 1450
–1454, 1998[Medline]
Ciancio G, Joshua M, Thomas A: Review of major clinical trials with mycophenolate mofetil in renal transplantation.
Transplantation 80[Suppl 2]
: S191
–S200, 2005
Shah S, Collett D, Johnson R, Thuraisingham RC, Raftery MJ, Rudge CJ, Yaqoob MM: Long term graft outcome with mycophenolate mofetil and azathioprine: A paired kidney analysis.
Transplantation 82
: 1634
–1639, 2006[CrossRef][Medline]
Allison AC, Eugui EM: Mechanisms of action of mycophenolate mofetil in preventing acute and chronic allograft rejection.
Transplantation 80[Suppl 2]
: S181
–S190, 2005
Weir MR, Ward MT, Blahut SA, Klassen DK, Cangro CB, Bartlett ST, Fink JC: Long-term impact of discontinued or reduced calcineurin inhibitor in patients with chronic allograft nephropathy.
Kidney Int 59
: 1567
–1573, 2001[CrossRef][Medline]
Dudley C, Pohanka E, Riad H, Dedochova J, Wijngaard P, Sutter C, Silva HT Jr; Mycophenolate Mofetil Creeping Creatinine Study Group: Mycophenolate mofetil substitution for cyclosporin A in renal transplant recipients with chronic progressive allograft dysfunction: The "creeping creatinine" study.
Transplantation 79
: 466
–475, 2005[CrossRef][Medline]
Merville P, Berge F, Deminiere C, Morel D, Chong G, Durand D, Rostaing L, Mourad G, Potaux L: Lower incidence of chronic allograft nephropathy at 1 year post-transplantation in patients treated with mycophenolate mofetil.
Am J Transplant 4
: 1769
–1775, 2004[CrossRef][Medline]
Halloran PF, Langone AJ, Helderman JH, Kaplan B: Assessing long-term nephron loss: Is it time to kick the CAN grading system?
Am J Transplant 4
: 1729
–1730, 2004[CrossRef][Medline]
Shah N, Al-Khoury S, Afzali B, Covic A, Roche A, Marsh J, Macdougall IC, Goldsmith DJ: Posttransplantation anemia in adult renal allograft recipients: Prevalence and predictors.
Transplantation 81
: 1112
–1118, 2006[CrossRef][Medline]
McNeil K, Glanville AR, Wahlers T, Knoop C, Speich R, Mamelok RD, Maurer J, Ives J, Corris PA: Comparison of mycophenolate mofetil and azathioprine for prevention of bronchiolitis obliterans syndrome in de novo lung transplant recipients.
Transplantation 81
: 998
–1003, 2006[CrossRef][Medline]
Wang K, Zhang H, Wei LQ, Yang Y, Lu Y: Safety of mycophenolate mofetil versus azathioprine in renal transplantation: A systematic review.
Transplant Proc 36
: 2068
–2070, 2004[CrossRef][Medline]
Vanrenterghem Y, Ponticelli C, Morales JM, Abramowicz D, Baboolal K, Eklund B, Kliem V, Legendre C, Morais Sarmento AL, Vincenti F: Prevalence and management of anemia in renal transplant recipients: A European survey.
Am J Transplant 3
: 835
–845, 2003[CrossRef][Medline]
Atasever A, Bacakoglu F, Toz H, Basoglu OK, Duman S, Basak K, Guzelant A, Sayiner A: Tuberculosis in renal transplant recipients on various immunosuppressive regimens.
Nephrol Dial Transplant 20
: 797
–802, 2005[Abstract/Free Full Text]
Kauffman HM, Cherikh WS, McBride MA, Cheng Y, Hanto DW: Post-transplant de novo malignancies in renal transplant recipients: The past and present.
Transpl Int 19
: 607
–620, 2006[CrossRef][Medline]
Related Article
This Month's Highlights
J. Am. Soc. Nephrol. 2007 18: 1617-1618.
[Full Text][PDF]
This article has been cited by other articles:
L. M. Ebah, N. Mapstone, S. P. Gibson, and A. Woywodt Diagnosis by inclusion
NDT Plus,
October 1, 2008;
1(5):
349 - 353.
[Full Text][PDF]
Top
Abstract
Introduction
75 kg) in combination with maintenance immunosuppressive therapy with the cyclosporine microemulsion Neoral and steroids (see Remuzzi et al. [15] for further details). From October 1997 to May 2001, 336 patients were included. However, two patients (one per group) did not receive the transplant for technical reasons. Thus, 334 patients (167 per group) received the study drugs (Figure 1). The randomization was centralized at the Laboratory of Biostatistics of the Clinical Research Centre for Rare Diseases of the Mario Negri Institute for Pharmacologic Research. At completion of the first 6-mo treatment period (phase A), steroid dosage was progressively tapered and discontinued in patients with (1) no more than two acute rejection episodes, (2) no steroid-resistant rejections, and (3) serum creatinine concentrations of 
177 µmol/L (Figure 1). When an acute rejection episode was diagnosed, oral steroid was renewed at the dosage before the last reduction or at 6 mg every other day if the patient had already discontinued the medication. Patients were then followed up to 21 mo after transplantation (phase B). 
2, or Fisher exact test as appropriate. Kaplan-Meier analysis was used to investigate whether survival functions differed between treatment groups. The difference between the survival curves was assessed by log-rank test. Cox proportional hazards models, with adjustment for main demographic and clinical covariates (site, donor and recipient gender, donor and recipient age, donor and recipient weight, and donor serum creatinine) were used to detect the effect of treatment on time-dependent end points. Proportional hazards assumptions were checked by means of Schoenfeld residuals. General linear mixed models were used to compare GFR levels at various time points (e.g., every 6 mo since transplantation, between regimen treatments). Statistical analyses were accomplished with SAS software (version 9.1.3; SAS Institute, Cary, NC). The statistical level of significance was P < 0.05. All statistical analyses were performed by two-tailed tests. The study was not initially powered to detect a difference in graft survival or function between treatment groups, and the recruitment was finalized to have all MYSS patients on active follow-up for at least 2 yr in the setting of the MYSS Follow-up study. 
